You are here

Using wrist pulse to understand body condition

A doctor once rightly said, “When we take the pulse, it is the soul of the doctor talking to the soul of the patient”. A pulse indeed contains some crucial pointers to derive at the right diagnosis. Just a reduced or increased pulse rate in itself is an indicator of certain ailments. Hence, the diagnostic potential of pulse measurement is immense. However, pulse measurement today is both an art and a science with multiple years of experience and skill required on the part of the physician. However with the advancement of modern science in recent times, there is a need to make a quantified diagnosis using wrist pulse. Computerized quantification of these pulse signals could add great value to our healthcare.

A heart beat generates pressure wave which propagate throughout the arterial system. The shapes of wrist pulse waveforms are influenced by their continuous interaction with the non-uniform arterial system. These waves expand the arterial walls as they travel along and the expansions are palpable as the wrist pulse. A typical pulse signal has a multi period trend. Systolic wave with higher amplitude contributes to the main component of the pulse signal. The diastolic wave contributes to the lower amplitude secondary wave of the pulse signal. The information regarding heart is contained in the systolic wave whereas the secondary wave provides information on the reflection sites and the periphery of the arterial system. Analyzing this information would help one detect abnormalities in the body condition. Many efforts have been made recently to analyze wrist pulse pressure signals using efficient computer based techniques

Professor Narayana Dutt and Rangaprakash who did this work at Indian Institute of Science (IISc), have tried to study the wrist pulse signals to distinguish changing body conditions. When a pulse is represented as a digital signal, it can be viewed as a wave with two maxima and three minima. Various combinations of differences between the times taken to reach these five points define the attributes of a pulse signal. It was observed that these attributes take different values when the body is under different conditions.

In their experiment, the researchers recorded the wrist pulse signals from several subjects in two recording conditions for two cases - before lunch and after lunch, before exercise and after exercise. After studying the change in the quantified attributes of the pulse signal in these conditions, it was observed that these attributes change significantly with the change in body conditions. The real value however, would be when these attributes tell us the body condition it corresponds to. The researchers tried to group these recorded signals based on their attribute values. They observed that the values fell into two distinguishable clusters for each case. In the lunch case, they had one cluster corresponding to the wrist pulse signals recorded before lunch and another cluster corresponding to the wrist pulse signals recorded after lunch. A similar result was observed in the case of before and after exercise. The attributes are found to be statistically significant with high classification accuracy of 99.71% for exercise case and 99.94% for lunch case, which clearly demonstrates the efficacy of the technique. This implied that the pulse signals recorded for a subject could tell us whether it was recorded before lunch or after and before exercise or after. Extrapolating this idea, we can imagine how a pulse signal could also indicate an abnormality in our body.

The highlight of Professor Narayana Dutt and Rangaprakash’s study is that it has demonstrated the feasibility of using the attributes of a wrist pulse signal to understand one’s body condition. There is a need for applying these results for various healthcare applications. Recording wrist pulse signals from patients in hospitals and then comparing it with signals from normal subjects could accomplish this. Hope the day is not too far, when your wristwatch or your favorite armband reads this signal of life and predicts an unseen abnormality - thus leading to a healthier society.

 

About the authors

D. Rangaprakash was with the Dayananda Sagar College of Engineering, Bangalore, and is currently at Auburn University, USA.

Prof Narayana Dutt is with the Department of Electrical Communication Engineering, Indian Institute of Science, Bangalore.

 

Contact: Narayana Dutt http://www.ece.iisc.ernet.in/~dndutt/DND.html

PHONE: +91 80 309 2283
EMAIL: dndutt@ece.iisc.ernet.in